SDXL

Note:To generate the best results, make the height and width atleast 896x896 size and steps minimum of 45 (the more details in prompt the higher the steps should be).Instruction for best results:On PC:Checkpoint - must be SDXL only or it won't work. Choose your best preference.Load Input/Base Face Image - best image quality equals best results.Pose Image Section - just upload a FULL BODY photo of a person if you want results that are not portrait.Steps - more steps gives more details, but it will take more credits. Suggested minimum steps is 15, it's up for you to decide. If your prompt has a lot of details better increase steps, I personally put it between 45 and 60.Height & Width - For best results, put it to 1080x1080.On mobile:First (1) upload image : Pose Image Section - just upload a FULL BODY photo of a person if you want results that are not portrait.Second (2) upload image : Load Input/Base Face Image - best image quality equals best results.I know this is more costly than the first face swap but I'm looking for the easiest way but cheaper workflow that gives great results.

INSTRUCTIONS (Click here to expand):IMPORTANT: Due to an update of the website that removed the LoRas strength numbers, it is recommended to adopt the following configuration: first lora (0.8), second lora (0.6) and third lora (0.8).This application seeks to simplify the creation of an image in anime style but at the same time achieve a good quality with high resolution and details.The dynamic is that you write what you want to create in the prompt spaces and the rest happens autonomously.You can also choose the number of images created simultaneously in the "Batch Size" option (justkeep in mind that the more images, the more credits it will cost).In addition, images are generated by XL format models with automatic upscaling to a resolution of 1152 x 1728 pixels.To run this app you don't need to select a specific VAE.If you want to use the App; go ahead, it would be very useful for me to test it.Enjoy, please 😁

Just write your prompt and generateenjoyBase on https://tensor.art/models/754415957910332809/CND_Beauty_SH_XL-ALPHA

Swap Your Face as Anime Character - Low Cost, High Quality

Create Realistic Video/Animations using your prompt. Enjoy. Please like and support if you like our AI Tools. :)

Hey guys , try out my tool and give feedback in commentsSUGGESTIONS :Try to match ur prompt with ref image style for best resultsUse proper lora to achieve best results, for anime u can use niji v6 massive style and ct-sdxl fantasy nijiUse Style strength between 0.6-1.0